Mohsen B. Mesgaran is an Assistant Professor of Weed Ecology at the Department of Plant Sciences, University of California, Davis. Prior to that, he worked as a research fellow and postdoc at the University of Melbourne and Stanford. He studies the eco-physio-evolutionary processes underpinning the spread and escalation of weedy and invasive plants both in agricultural and natural systems. He uses a diverse variety of tools and techniques including advanced modeling, remote and proximal sensing, image processing, and GIS to develop “eco-tech” solutions for ecological management of weeds as well as a better understating of the spatial and temporal variability of weed populations and communities.
Title: A deeper look at weeds with deep learning
Recent advances in machine vision and sensing technology coupled with the power of high-performance computers have enabled us to both generate and analyze massive amount of data. Data analytic tool such as machine learning has been revolutionary in this rapidly evolving area and particularly holds promise in precision (digital) agriculture. Deep learning, a subfield of machine learning, has shown human-level or even outperformed human’s eyes in detecting objects and patterns in various image-based problems. In weed research and management deep learning offers a new set of valuable implications from accurate discrimination of weeds from the crop and automated weed identification to examining the seed viability or seed classification. In this presentation, I will provide an overview of deep learning technique and its uses in studying weeds.